"EDDTableAggregateRows" - Time, Latitude and Longitude min/max values are calculated from the first child only

[MarcoAlbaETT]

Describe the bug
EDDTableAggregateRows range for time is not the min/max of all the child.
Also other variables, like latitude and longitude have this problem.
Some variables are not affected by this problem and the min/max value is calculated from all the child

To Reproduce

1. Create a "EDDTableAggregateRows" datasets with two (or more) child of type ""EDDTableFromErddap"
2. Time of the first datasets must be different from the other. As an example, first child has time from 2013-01-01 to 2020-03-31 and the second one has data from 2024-01-01 to today
3. The "EDDTableAggregateRows" time will be from 2013-01-01 to 2020-03-31

With real data, if you create this dataset:

<dataset type="EDDTableAggregateRows" datasetID="TS_SLEV_TAD" active="true">
	<reloadEveryNMinutes>60</reloadEveryNMinutes>
	<dataset type="EDDTableFromErddap" datasetID="TS_SLEV_TAD_15_composite" active="true" >
		<reloadEveryNMinutes>60</reloadEveryNMinutes>
		<sourceUrl>https://er1webapps.emodnet-physics.eu/erddap/tabledap/TS_SLEV_TAD_15</sourceUrl>
	</dataset>
	<dataset type="EDDTableFromErddap" datasetID="TS_SLEV_TAD_800_composite" active="true" >
		<reloadEveryNMinutes>60</reloadEveryNMinutes>
		<sourceUrl>https://er1webapps.emodnet-physics.eu/erddap/tabledap/TS_SLEV_TAD_800</sourceUrl>
	</dataset>
</dataset>

https://er1webapps.emodnet-physics.eu/erddap/tabledap/TS_SLEV_TAD_15 time is 2013-06-07T09:00:00Z to 2013-06-13T09:57:15Z
https://er1webapps.emodnet-physics.eu/erddap/tabledap/TS_SLEV_TAD_800 time is 2022-09-01T00:03:00Z to 2023-12-19T11:03:00Z
The EDDTableAggregateRows TS_SLEV_TAD time is 2013-06-07T09:00:00Z to 2013-06-13T09:57:15Z (instead of 2013-06-07T09:00:00Z to 2023-12-19T11:03:00Z)
Same problem with latitude and longitude.
Other variable (like SLEV) min/max are calculated in the right way.

Expected behavior
time range, latitude and longitude of the EDDTableAggregateRows dataset to be the min / max of all the child datasets

Screenshots
not applicable

Server
I'm using ERDDAP Version 2.23 in a docker container using axiom/docker-erddap:2.23-jdk17-openjdk image

Desktop (please complete the following information):

• Windows, Linux
• Any Browser

Additional context
None

[MarcoAlbaETT]

Hi!
I've (probably) found the problem in the code for the time variable in the EDVTimeStamp class.
I'll make some test and if all is fine I'll make a pull request with the changes.

I'm still working on the same problem for latitude, longitude and depth variables.

[ChrisJohnNOAA]

@MarcoAlbaETT That's great. I've assigned this issue to you since you're working on it. Let me know if you have any questions.

[MarcoAlbaETT]

Hi @ChrisJohnNOAA, the bug fix is ready.
Sadly I've trouble running the tests. I've changed only EDDTableAggregateRows.java class, but the tests fail for EDVTimeStampTests. I think this is something related to the language settings of my pc.
Is it fine to make a pull request without a succesful test?

[ChrisJohnNOAA]

Yeah, go ahead and make the pull request. I'm aware that some tests can fail on other machines and have been trying to find and fix those to make them more reliable.